AiCostCompare

Compare models

Pick up to 4 models for a side-by-side spec sheet. Hover any dotted metric name for what the test measures. Best value in each row is highlighted.

Claude Sonnet 5Mercury 2
Claude Sonnet 5

Anthropic

Mercury 2

Inception

Pricing
Input $/1M$2.00$0.25
Output $/1M$10.00$0.75
Blended $/1M (3:1)$4.00$0.375
Cache read $/1M$0.20$0.025
Cache write $/1M$2.50
Quality indices
Intelligence Index38.213.8
Coding Index71.531.1
Math Index
Agentic Index43.61.6
Speed & latency
Output speed (tok/s)82607
Time to first token118.14s7.16s
Time to first answer token118.14s7.16s
Benchmarks (Artificial Analysis)
GPQA Diamond91.1%77.0%
Humanity's Last Exam41.3%17.1%
SciCode54.3%37.7%
IFBench69.8%
AA-LCR (Long Context Reasoning)82.0%43.7%
Terminal-Bench Hard26.5%
Terminal-Bench 2.180.5%27.3%
τ²-Bench (Telecom)70.8%
τ³-Bench Banking37.3%9.5%
Design Arena (Elo)
Website1,2881,038
Web apps1,260
Full-stack1,260
Mobile apps1,249
Android native1,240
UI components1,299998
Data viz1,2601,015
SVG1,2121,016
3D1,2891,001
Game dev1,3131,008
Agentic game dev1,227
Godot game dev1,238
HTML slides1,220
Python→PPTX slides1,216
Code categories1,2931,028
ASCII art1,2261,021
Specs
Context window1M128K
Max output tokens128K50K
Input modalitiestext, image, filetext
TokenizerClaudeOther